UniNest: Housing Price Estimator

Medium Artcile:

Data geared project designed to project housing market prices in University cities

Web Scraping

Created a web scraper from scratch using Python and HTML-Requests to scrape the following from over 1200 listings across 6 cities:

Location
Title
Price
Style
Bedrooms
Bathrooms
Size
Air Conditioning
Den
Bachelor/Studio

Data Cleaning

Cleaned the raw data scraped from Kijiji to use for EDA, graph plotting, and model building. The cleaning included:

Removing "Wanted" listings
Streamlined bedrooms, bathroom and price to include integers only
Parsed through bedrooms to create binary columns for dens
Removed "Please Contact" listings

Exploratory Data Analysis

Analyzed the clean data and making bar, scatterplot and heatmap graphs using Tableau & Jupyter Notebook [Seaborn, Matplotlib]

The following graphs analyze relationships between the various data columns including Size, Price, Bathrooms, Bedrooms, Location and Den status:

View the Dashboard: https://public.tableau.com/views/HousingPriceBoard/Dashboard1?:language=en-US&publish=yes&:display_count=n&:origin=viz_share_link

Machine Learning Model

Trained a multiple linear regression model with sklearn (accuracy ~70-80%) to predict housing data based off the information I scraped and cleaned. Deployed the model using a Flask API.

Resources & Packages Used

Python Version: 3.11.4

Packages: pandas, numpy, matplotlib, seaborn, html-requests